251 The Production of Anthocyanins and Anthocyanidins.\#151 ; Part III.* By Arthur Ernest Everest , D.Sc .
, Ph. D. ( Communicated by Dr. F. W. Keeble , F.R.S. Received April 9 , 1918 .
) The work of Willstatter , f of Willstatter and Mallison , * and of Willstatter and Zechmeister , S that has followed upon the proposal by the present author of a structural formula for the anthocyan pigments , || has clearly demonstrated the correctness of the formula proposed , and of the conclusions of the present author as to the relationship that exists between the anthocyan and flavonol groups , exemplified thus :\#151 ; Cl HO\#151 ; OH C\#151 ; / \oH II X--OH C\#151 ; OH I c HO || O O OH HO\#151 ; '/ OH n Q0 OH \A/ C_0H I c HO | H Flavonol ( Myricetin ) .
Anthocyan ( Delphinidin ) .
Although it has thus been proved that anthocyanidins can be produced from flavonols , and that the flavonol glucosides pass , likewise , to anthocyanins , it does not of necessity follow that the anthocyan pigments are thus formed in plants and flowers , although this is probable , and in the present paper chemical evidence is produced which strengthens this probability .
It was with a view to an examination of some points that may prove of value in connection with the tracing of the formation of the anthocyan pigments in Nature that the author began the investigations , of which this is a preliminary announcement .
In view of the fact that the work , which was begun during the summer of 1914 , has had of necessity to stand aside , and must continue to do so for some time , and of the publication of further work by Willstatter and his collaborators , IT some of which touches upon the same ground , the author feels that the publication of the present preliminary observations will not be out of place .
* Parts I and II appeared in these ' Proceedings , ' B , vol. 87 , p. 444 ; vol. 88 , p. 326 ( 1914 ) .
+ ' Sitzber .
K. Preuss .
Akad .
Wiss .
, ' 1914 , p. 402 .
\ Ibid. , p. 769 .
S Ibid. , p. 886 .
II *Boy .
Soc. Proc. , ' B , vol. 87 , p. 444 ( 1914 ) .
H Willstatter and Weil , ' Annalen , ' vol. 412 , p. 178 ( 1916 ) .
X 2 Dr. A. E. Everest .
Both Willstatter and the author have noticed that when fiavonols are reduced as described by them , with formation of anthocyans , a considerable proportion of the yellow compound remains unacted upon\#151 ; as the result of quantitative estimation Willstatter found , in one case , only 4 per cent , of anthocyan present after reduction\#151 ; but this is , to some extent at least , due to the difficulty in regulating the reduction of the pigment , and it is quite conceivable that in Nature conditions may be more favourable to anthocyan formation than in the experiments thus far described .
It would appear desirable that evidence should be collected upon which conclusions might be drawn as to whether the anthocyan pigments are produced in plants vid flavonols , or , as the result of direct synthesis , independent of the presence of flavonols .
If they are produced vid flavonols , then , unless the conditions present in Nature allow the reaction causing the conversion of the flavonol into anthocyan to proceed to completion , and the reaction proceeds more rapidly than that whereby the flavonol is synthesised , and , further , that there is no reconversion from anthocyan to flavonol as the result of re-oxidation , it is to be anticipated that flowers , fruit , etc. , containing an anthocyan pigment will also contain some quantity , great or small , of a flavonol , and indeed of the particular flavonol that corresponds to the anthocyan present .
Thus , where the anthocyan pigment is one of the cyanidin glucosides , a corresponding quercetin glucoside would be expected , whereas if the anthocyan were a glucoside of pelargonidin , then a corresponding glucoside of kaempferol should be found , and so forth .
It is , indeed , possible that if these pigments are produced vid flavonols , there is a more or less regular relationship existing between the amount of flavonol and of the corresponding anthocyan present in a normal full-grown flower of any particular species when grown under normal conditions .
There remains a further possibility which should not be overlooked , viz. , that where two flavonols are present under conditions which give rise to production of anthocyans from flavonols , one of the yellow pigments may be preferentially reduced , passing completely to an anthocyan before the second is attacked .
This seems , at first sight , rather unlikely , but if such were the case , the result would be that not a flavonol corresponding to the anthocyan would be found , but one differing from it ; thus the investigations proposed by the author would produce useful evidence in this case also , and , if necessary , could probably be verified by independent methods .
With respect to the presence of yellow sap-pigments , flavones or flavonols , in flowers containing anthocyans , there is considerable evidence of their co-existence already available , and many of the early workers noted that the The Production of .
Anthocyanins and Anthocyanidins .
253 anthocyan pigments were accompanied by yellow sap-pigments.* During the preliminary work on the cornflower pigment a yellow sap-pigment was isolated in small quantities by the author , though the product was not investigated , nor was the fact recorded in the published results , f whilst Wheldale has stated that in various varieties of Antirrhinum , anthocyans are present together with flavones , but in none of these cases have both anthocyan and flavone , or flavonol , from any one flower been isolated and identified.^ Shortly after publication of his conclusions concerning the formation of anthocyans from yellow sap-pigments , the author began preliminary work with a view to ascertaining , in a systematic way , the identity of both anthocyan and yellow sap-pigment present in various flowers , by isolating both of these from the same batch of petals , and for this purpose , in the first instance , a series of self-coloured Violas was chosen .
Apart from the fact that the presence of yellow sap-pigments in flowers which contain anthocyans has been thus observed , there is but little evidence yet available as to the actual structure of both flavone ( or flavonol ) and anthocyan pigments present in any one flower .
It would appear that the case in which the published data concerning these two types of pigment most nearly approach that desired is that of Delphinium consolida , L. A. G. Perkin and Wilkinson , S following up the observations of A. G. Perkin and Pilgrim , || isolated kaempferol o RO-f'y ' ^C\#151 ; / N\#151 ; OH .
II N\#151 ; x \A/ '-oh I C .
HO || O ( they failed to isolate the glucoside , in which form the colour was originally present in the flowers ) from the blue flowers of this plant , both when they used the flowers , as also when they made use of an extract purchased from Merck , of Darmstadt .
Willstatter and Mieg,1T who examined the anthocyan pigment present in Delphinium consolida , L. , using the commercial purple " Flores * * * S * Hope , ' Journ. Prakt .
Chem. , ' 1837 ( 10 ) , p. 269 ; Fremy and Cloez , ' Journ. de Chim .
et de Phys. ' [ 3 ] , vol. 25 , p. 249 ; Filhol , 'Compt .
Rend .
, ' vol. 39 , p. 194 ; 'Journ .
Prakt .
Chem. , ' vol. 63 , p. 78 ( 1854 ) ; Martens , ' Jahrber .
, ' 1855 , p. 657 .
+ Willstatter and Everest , * Annalen , ' vol. 401 , p. 189 ( 1913 ) .
t Cf .
further Willstatter and Mallison , ' Annalen , ' vol. 408 , p. 157 ( 1915 ) .
S 'Journ .
Chem. So .
, ' 1902 , p. 585 .
'Journ .
Chem. Soc. , ' 1898 , p. 267 .
IT 1 Annalen , ' vol. 408 , p. 61 ( 1915 ) .
Dr. A. E. Everest .
Calcafcrippae , " obtained from these flowers a complex diglucoside anthocyan , to which they gave the name delphinin , which , on hydrolysis , yielded not the anthocyan that would correspond to kaempferol ( pelargonidin ) , but that corresponding to myricetin , viz. , delphinidin .
Cl O ____OH HO\#151 ; NoH I L ft OT OH \/ V/ c-OH l c HO | H In view of the fact that the colour descriptions of the flowers used do not coincide , * and that Willstatter and Mieg ( .
state that other varieties of Delphinium appear to contain colouring matters which differ , at least , from their glucoside delphinin , it is not certain whether the results of these researches can be safely used as evidence that in Delphinium consolida , L. , the anthocyan present is not that which corresponds to the flavonol derivative ( yellow sap-pigment ) that is also present .
Wheldalef has somewhat fully investigated the Mendelian factors in the Antirrhinum , and , in the course of her work , has isolated apigenin and luteolin from ivory and yellow varieties respectively , and extracted anthocyan pigments from various red-coloured flowers of the same plant .
Here the evidence is insufficient , as Wheldale has not as yet obtained the anthocyan pigments in a crystalline condition nor identified their structure ; moreover , although she assumes that certain of the red flowers contain apigenin and luteolin in addition to the anthocyan pigments , it does not appear that she has isolated the flavone derivatives actually from the red flowers , but only from ivory or yellow specimens .
It is to be hoped that Wheldale will continue her work on these products , particularly in view of the fact that , if the anthocyan pigments are produced in Nature vid flavone , or flavonol , derivatives , and her assumption that the red flowers contain apigenin and luteolin , as do the ivory and yellow , proves correct , then the anthocyan pigments she has under investigation should be members of a new series which are related to flavone derivatives , not , as are all the anthocyans thus far described , to flavonols .
The impossibility of assuming that , because a particular pigment is present in the flowers of one variety of a plant it will occur in those of other * The author having submitted a sample of the purple Delphinium used by Willstatter and Mieg , to Prof. A. G. Perkin , he kindly confirmed the fact that it did not coincide with the product used by him .
t ' Journ. Genetics , ' vol. 4 , p. 109 ( 1914 ) ; and earlier papers referred to therein .
The Production of Anthocyanins and Anthocyanidins .
255 varieties of the same species , will be fully realised when the work of Willstatter and his collaborators on the anthocyan pigments , in particular Willstatter and Mallison , * or that of A. G. Perkin , * on the various cotton flowers , f is considered .
One further case needs consideration : CombesJ states that he has isolated a crystalline anthocyan pigment from the red autumn leaves of Ampelopds hederacea , whilst from green summer leaves of the same plant he obtained a yellow-brown sap-pigment in crystalline condition .
He further states that , by reduction of the yellow compound , he obtained a substance ( crystalline ) that was identical with the anthocyan , and vice versed ; by oxidation of the anthocyan he obtained the yellow-brown compound in crystalline form , and identical with that obtained from the green leaves .
In this case neither anthocyan nor yellow colouring matter has been identified , whilst the published data , which are scanty , leave the impression that his anthocyan may consist of crystals of a flavone derivative coloured red .
The Colouring Matters of the Viola .
At the time when the present work was commenced , our knowledge of the colouring matters of the Viola species was confined to the yellow pigments .
Viola-quercitrin , from Viola tricolor , had been isolated and investigated by various workers , and finally proved to be quercetinrhamno-glueoside , S whilst A. G. Perkin 11 had shown that in the ordinary Violet ( odor aid ) there is also present a glucoside of quercetin .
It is interesting to note that the latter investigation was carried out with the ordinary violet-coloured flowers , not with white Violets ( this was not specifically stated in the paper cited , but has been intimated to the author , by Prof. Perkin).IT The author proposes to examine systematically the different species of Viola , with a view to isolation and identification of both the yellow flavone , also the anthocyan pigments present in them .
At the outset , self-coloured varieties were chosen , and , feeling that the anthocyan would probably give more trouble in isolation than the yellow pigment , one that contained a very large percentage of anthocyan pigment\#151 ; Sutton 's " Black Knight " \#151 ; was first taken for investigation .
The plants from which the flowers were gathered were mostly grown for * ' Annalen , ' vol. 408 , p. 147 ( 1915 ) .
+ ' Jouru .
Chem. Soc. , ' 1916 , p. 145 .
X ' Count .
Rend .
, ' vol. 157 , p. 1002 ; vol. 157 , p. 1454 .
S A. G. Perkin , 'Journ .
Chem. Soc. , ' 1910 , p. 1776 .
|| ' Journ. Chem. Soc. , ' 1904 , p. 58 .
The author proposes to isolate and identify the anthocyan pigment from the Violet , \lt ; and thus complete the data .
Dr. A. E. Everest .
the author by Mr. Costin , of the Horticultural Department , University College , Reading , to whom , as also to Messrs. Sutton and Sons , Reading , for their interest and help , the author desires to express his thanks .
The flowers were an almost dead black colour , and , when viewed by Strong transmitted light , a deep purple to purple-black .
The petals only were gathered , and these were air-dried , in the shade , at room-temperature , then completely dried in vacuo [ over cone , sulphuric acid and caustic soda ( solid ) ] , powdered , and stored for use .
Contrary to anticipation , the isolation and purification of the anthocyan pigment gave but little trouble , but the method used by the author prior to the publication of the paper on the blue-black Pansy pigment by Willstatter and Weil ( loc. cit. ) was not quite so expeditious as that described by those authors .
Although Willstatter and Weil described their flowers as being blue-black , a description that could not hold for the decidedly purple-black flowers used by the present author , the data given by them concerning the pigment they obtained made it appear most probable that it was identical with that obtained by the present author .
The author anticipated obtaining a cyanidin derivative , but , having isolated a quantity of the pigment , he was prevented on account of lack of time from proceeding to the identification of the product and to the completion of the isolation and identification of the yellow pigment that had been shown to be present , and proved capable of giving rise to anthocyan by the author 's reduction method , as this involved a considerable undertake ing .
The timely appearance of the paper by Willstatter and Weil , in which it was shown that they had obtained a delphinidin derivative , of which very clear and characteristic description was given , and which further contained data concerning delphinidin itself that greatly facilitates its identification , has enabled the author to show , with but little work and with the small quantity of pigment he had available , that his pigment is undoubtedly a derivative of delphinidin , not of cyanidin , and is without doubt identical with Willstatter and Weil 's violanin .
It is entirely due to the above publication that the author has been able to bring together his observations in their present form , for beyond the fact that it enabled the author so readily to identify his anthocyan pigment as a delphinidin derivative , this identification made possible some conclusions as to the yellow sap-pigment also .
If the yellow sap-pigment present is that which corresponds to the anthocyan , then , in this case , it should be a myricetin glucoside .
Myricetin differs from all the other known flavonol and flavone sap-pigments in that it dissolves in dilute alkalis forming a green solution , which passes on exposure The Production of Anthocyanins and 257 to air to blue , violet , then orange yellow ; * moreover , in the presence of other yellow sap-pigments , a green colour passing to a brown-yellow is observed when my rice tin is present .
In this way its presence , together with quercetin , is shown in the colouring matter obtained from occidentalis .
The known glucoside of myricetin , on the other hand , gives only a yellow coloration with dilute alkalis .
How whilst the identity of quercetin , or any similar sap-pigment of the flavonol group , would have to be established by isolation and characterisation , pending time for further investigation , the presence of myricetin would be almost conclusively proved if the above-mentioned characteristic colour changes could be observed under proper conditions .
From the observations made in the course of the present investigations , and described below , there can now be no doubt that , unless gossypetin is present , which is unlikely , there is present in the Yiola " Black Knight " ( Sutton 's ) , a glucoside of myricetin side by side with a glucoside of delphinidin .
Thus the presence in one flower of an anthocyan pigment and of the flavonol derivatives from which it would be produced by reduction , is shown for the first time .
Experimental .
Isolation of the Anthocyan Pigment of the Purple-Black Viola .
The dry powdered petals of Sutton 's " Black Knight " ( 25 grm. ) were extracted with alcoholic hydrochloric acid ( 200 c.c. of a mixture made up in the proportion 95 c.c. of 96-per-cent .
EtOH ; 5 c.c. of 2N aqueous HC1 ; and 1 c.c. cone .
aq .
HC1 ) , and after standing some hours the extract filtered off and the residue further extracted with the same solvent .
( As stated by Willstatter and Weil , MeOH may , with advantage , replace EtOH .
) The combined filtrates were mixed\#151 ; good agitation\#151 ; with 2\#151 ; 2\#163 ; times their volume of ether , whereby most of the anthocyan pigment was precipitated as a sticky mass from which the liquors were readily decanted ; the liquors were then filtered and worked up for the yellow pigment which they contained .
Prior to the publication of Willstatter and Weil 's paper on the blue-black Pansy ( joe .
cit. ) , the crude anthocyan pigment\#151 ; after one reprecipitation from alcoholic hydrochloric acid by means of ether\#151 ; was fractionally precipitated from the same solvent by ether , then finally crystallised by solution in acidified alcohol , addition of aqueous acid ( HC1 ) to the solution , and slow evaporation of the alcohol , as described by Willstatter and Everest , * A. G. Perkins , ' Journ. Chem. Soc. , ' 1899 , p. 1289 .
Dr. A. E. Everest .
for the purification of cyanin chloride from the Cornflower.* As Willstatter and Weil 's method of purification would save considerable time , it was tested with the above crude product , and found to be very satisfactory , viz. : The moist precipitate of the anthocyan pigment , either crude or preferably after one reprecipitation from alcoholic hydrochloric acid by means of ether , was dissolved in 0'04-per-cent , aqueous HC1 ( for the total precipitate from 25 grm. dry petals ca .
50 c.c. were used ) , the solution filtered , and the filtrate mixed with half volume of 96-per-cent , ethyl alcohol , and one volume of 4-per-cent , hydrochloric acid , then allowed to stand several days in a dish loosely covered with filter paper .
Under these conditions well-formed crystals separated which were collected , washed with 7-per-cent , aqueous hydrochloric acid , and air dried ( 25 grm. petals gave 1*3 grm. crystals ) .
The pigment when recrystallised in this way , separated with splendid regularity as deep red , clean cut , hexagonal or tetrahedral plates .
The properties of the crystalline chloride have been carefully noted and compared with those of Willstatter and Weil 's violanin chloride , and the result of the investigation leaves no , doubt that the products are identical .
In view7 of the discovery in the alcohol-ether liquors of the presence not only of myricetin , but also of another flavone derivative which is capable of producing an anthocyan by reduction , an attempt was made to see whether any indication of the presence of other anthocyanins could be detected in the plant extract .
To this end , the alcoholic hydrochloric acid extract from 1 grm. of petals was carefully fractionally precipitated by ether , the precipitate from each fraction being collected , dissolved in water , made very faintly acid , then tested in respect of the colour reaction given with ferric chloride .
In all , nine fractions , which included the whole of the red pigment present , were made , all were treated in exactly the same way , and direct comparison of the colour reactions given by all nine fractions was made at the same time .
Ho difference of any kind could be observed in the colours produced , which in every case were clear pure blue .
This points to the absence of the cyanidin glucosides , cyanin , mekocyanin , idsein , asterin , and chrysanthemin , also to the absence of the glucosi , de derivatives of delphin-idin , myrtillin , althein , the anthocyanin from Vitis and possibly ampelopsin , for all of these pigments produce a violet coloration under the conditions of the experiment described above , but does not preclude the presence of other anthocyanins which either give blue , or violet-blue , or do not give any coloration under the above conditions .
The quantities of material available were not sufficient to allow of further investigations on these lines .
# * ' Annalen , ' vol. 401 , p. 189 ( 1913 ) .
The Production of Anthocyanins and Anthocyanidins .
259 In the comparison of the author 's anthocyanin chloride with Willstatter and Weil 's violanin chloride , the following points were examined :\#151 ; crystalline form , colour and reflex of the crystalline chloride ; colour of solution in aqueous acid and in alcohol ; colour change on addition of ferric chloride to an aqueous , or alcoholic , solution ; colour change on addition of alum to an aqueous solution ; solubility in acidified methyl alcohol , in acidified ethyl alcohol , in water , in aqueous hydrochloric acid ( 0*15 , 0*3 , 0*5 , 5 .
20 , and 30 per cent. HC1 respectively ) , in aqueous sulphuric acid ( 0*05 and 7 per cent ) : the distribution of pigment between amyl alcohol and dilute aqueous acid ; and finally the crystalline form , colour , and solubility in water of the picrate .
A determination of the water of crystallisation carried out by Mr. A. J. Hall , B.Sc. , A.I.C. , showed that when dried to constant weight at the temperature of a boiling water-bath and pressure 10 mm. , the air-dried crystalline chloride lost 15*68 per cent , of its weight ( used 0*5658 grm. air-dried product ; loss \#151 ; 0*0887 grm. ) .
Willstatter and Weil found that violanin chloride ( air-dried ) when dried in vacuo over sulphuric acid lost 15*29 and 16*15 per cent. ( 2 separate estimations ) .
Whilst the author found very slight divergences from Willstatter and Weil 's descriptions in respect of colour and reflex of the crystalline chloride , which the author would describe as red-brown and golden respectively , and the solubility in very dilute ( 0*05 per cent. ) sulphuric acid , in which it is rather more soluble than the description of Willstatter and Weil would lead one to assume , the solubility being nevertheless very small , in every other point , save the alum reaction , which was very pronouncedly different , the pigments possessed identical properties , and there can be no doubt that the pigment isolated by the author is violanin chloride .
The striking difference found in the alum reaction , which at first led the author to believe that his product was different from violanin chloride , has been satisfactorily explained , as will be seen from the results below .
The Alum Reaction of Pigments .
In a number of instances Willstatter and his collaborators have included among the characteristic reactions of anthocyan pigments described by them , beyond a colour reaction with ferric chloride , one also with alum or , as in the case of malvin and malvidin , they have stated that pigments give neither colour reaction with ferric chloride , nor with alum .
In the case of violanin chloride , they state that alum gives a blue coloration , whereas when the author examined the pigment obtained by him from the purple-black Viola , and which is doubtless identical with violanin chloride , he was unable to obtain any such coloration .
Dr. A. E. Everest .
Noting that the colour reactions of the anthocyans described by Willstatter as resulting from the addition of alum in aqueous solution were always the same as those produced by ferric chloride in aqueous solution , it occurred to the author that the reactions obtained by Willstatter with alum may possibly have been in reality the iron reaction , due to the use of impure alum containing traces of iron , whilst the author 's failure to obtain the reaction with his pigment was due to the fact that , on account of the knowledge that a ferric chloride reaction was given by the pigment , an alum free from iron had been obtained for the test .
To test this , a sample of commercial alum was obtained and used , whereupon the reaction was obtained exactly as anticipated , and as described by Willstatter .
The presence of an exceedingly small trace of iron in the sample of alum originally used by the author was shown by adding a very considerable bulk of it to a small quantity of the solution of the pigment , when the colour reaction slowly appeared ( No. 3 , below ) .
Reagent .
When added to solution of the crystalline chloride in faintly acid aqueous solution , produced 1 .
Dilute solution of pure alum ... ... ... ... .
2 .
Saturated solution of pure alum . .
* ... . .
3 .
Saturated solution of pure alum with addition of considerable quantity of excess solid 4 .
Dilute solution ( same concentration as No. 1 ) of commercial alum 5 .
Dilute ferric chloride ( aqueous ) ca .
6 per cent , solution , \ drop 6 .
Very dilute aqueous ferric chloride , 0*1 c.c. , containing 0*000015 grm. ferric chloride 6a .
Dilute aqueous solution of ferric sulphate , 2 drops 7 .
Dilute ferric chloride , as No. 5 ... ... . .
8 .
Dilute aqueous solution of pure alum ... . .
9 .
Dilute aqueous solution of commercial alum 10 .
Dilute aqueous solution of ferric sulphate ( same solution as No. 6a ) , 2 drops 11 .
Solid , finely divided ferric sulphate ... 12 .
Test No. 11 on dilution with water ... ... No colour change .
Slight and very slow change towards purple .
Slowly purple , then pure blue .
Very soon purple , then pure blue .
Immediately through purple to pure blue .
Very soon purple , then pure blue .
Very soon purple , then pure blue .
When added to solution of the pigment in faintly acid ethyl alcohol\#151 ; Immediately through purple to pure intense blue .
No colour change .
No colour change .
No colour change .
No colour change , even on standing , only slight change of tint towards purple .
Rapidly passes through purple to fine pure blue .
The fact that the addition of alum , containing iron , to an alcoholic solution of the colouring matter fails to give the reaction is thus shown to be due to the iron being in the form of the sulphate .
The Production of Anthocyanins and Anthocyanidins .
261 It would appear , therefore , that the colour reactions of anthocyans described by Willstatter and his collaborators as due to alum* are due to the presence of iron in their reagent , and not to the action of pure alum .
Hydrolysis of the Anthocyanin ( ) Chloride .
As the hydrolysis of the anthocyanin chloride had not been carried out prior to the appearance of Willstatter and Weil 's paper on the blue-black Pansy , the method described by them for the hydrolysis of violanin chloride was adopted .
By carrying out the hydrolysis quantitatively , the author found that 0*4730 grm. of his anthocyanin chloride ( violanin chloride ) , dried as described above , yielded 0*2205 grm. of anthocyanidin chloride ( delphinidin chloride ) , dried at 100 ' C. ; this included the small quantity recovered from the filtrate by means of amyl alcohol , the anthocyanidin chloride == 46*61 per cent , of the weight of dry anthocyanin chloride taken .
Willstatter and Weil found , for violanin chloride , a yield of 44 and 43 per per cent , ( two experiments ) of delphinidin chloride\#151 ; they did not recover the small portion of pigment present in the filtrate\#151 ; and state that theory requires 53*85 per cent , if violanin chloride is C27H29O16CI , whereas if it is O27H31O16CI , it should yield 52*37 per cent. The exact formula is still uncertain ( cf. Willstatter and Weil , loc. cit. ) .
The anthocyanidin chloride obtained by the author separated in minute micro-crystalline aggregates , not .
in the well-defined crystals obtained in the case of cyanidin chloride .
For identification purposes , the author 's hydrolysis product was tested in respect of the following characteristics , viz. , solubility in methyl alcohol , in ethyl alcohol , in water , in dilute sulphuric acid , and in hydrochloric acid ( various concentrations ) ; extraction from aqueous solution by ether , and removal of trace of pigment extracted by shaking with dilute acid ; ferric chloride reaction , in aqueous solution , in alcoholic solution , and the effect of dilution with water in the latter case upon the coloration produced ; the colour change produced when sodium carbonate is added to an acid solution ; and , finally , the effect of heat upon the dry product .
In every case it was found that the results obtained exactly coincided with those obtained by Willstatter and Weil ( loc. cit. ) or Willstatter and Miegf for delphinidin chloride .
The deposition of the violet colour base from the solution in water first formed by the chloride , and the absence of such deposition whep * U.g. , cf. ' Annalen , ' vol. 408 , pp. 35 , 38 , 50 , 130 , 133 , 141 ( 1915 ) .
+ ' Annalen , ' vol. 408 , p. 61 ( 1915 ) .
Dr : A. E. Everest .
warm water is added to a solution of the chloride in alcohol , which characterise delphinidin , were readily observed with the author 's product .
In order to confirm further the identity of the substance , a portion was converted into the picrate , which separated as bundles of fine red-brown needles , which proved to be difficultly soluble in water ; this agrees with Willstatter and Mieg 's description of delphinidin picrate .
A further portion was treated exactly as specified by Willstatter and Weil for the formation of their third hydrate of delphinidin chloride , and , under these conditions , the author 's product readily separated in the form of beautiful brown needles , singly or in clusters , which coincided exactly with the description and illustration given by Willstatter and Weil for the third hydrate of delphinidin chloride ( CisHnOyClAHaO ) .
When another portion was treated as described by those authors for the production of their fourth hydrate of delphinidin chloride ( Ci5Hu07C1,1^H20 ) , the changes described by them were clearly observed , viz. , immediate precipitation of the amorphous product , followed by gradual change , on standing , from the amorphous state into small well-defined crystals .
In this case , however , the author found it necessary , to keep the product somewhat above room-temperature , after the addition of the concentrated acid , in order to obtain the change from amorphous to crystalline state .
There can be no doubt that the author 's anthocyanidin is identical with delphinidin .
Examination of the Yellow Sap-Pigment .
The ether-alcohol liquors from the precipitation of the anthocyan pigment were shaken with dry , precipitated calcium carbonate ( BaCC\gt ; 3 was found less satisfactory ) , stood for about five minutes , filtered , and the ether , then alcohol , distilled off from the liquors thus obtained .
In order to discover the colour produced by the action of alkalis on the flavone glucosides present in this concentrate , a test portion was evaporated to dryness , extracted with ether , the ethereal extract ( which contains waxes in addition to flavone derivatives ) washed quite free from anthocyans by means of acid , then filtered , and finally shaken with dilute sodium carbonate solution , when a clear yellow aqueous layer was formed .
The bulk of the concentrate produced by removal of the ether and alcohol from the liquors obtained from the precipitation of the anthocyan was boiled with sulphuric or hydrochloric acid , whereby the glucosides were hydrolysed , the product cooled and extracted with ether , into which the waxes and flavone derivatives passed , whilst the anthocyanidin remained almost quantitatively in the aqueous acid .
The ether layer thus obtained , The Production of Anthocyanins and Anthocyanidins .
263 after separation , was washed six times with fresh quantities of aqueous acid of varying concentrations to remove all traces of anthocyan .
The last four washings were quite devoid of colour , and , on making alkaline with sodium carbonate , gave but the faintest tinge of yellow .
The ether layer , after receiving the above treatment , was filtered through double filter papers to remove all traces* of suspended plant matter , then shaken with dilute sodium carbonate solution , whereupon a fine deep , clear olive-green aqueous layer was produced .
On standing , this colour soon passed through a rather bluer shade of green to a brown-green , then to orange-yellow .
These observations , together with the fact that the glucoside gave yellow , not green , with dilute alkalis , gives a strong indication of the possible presence of myricetin .
Further confirmation of the presence of a myricetin ( or less probably gossypetin ) glucoside in the Violas under investigation was obtained in the following way:\#151 ; When in the original precipitation of the anthocyan pigment the extract obtained from the flower petals was precipitated by means of the addition of 3| times its volume of ether instead of 2| , the process above described gave an ether-alcohol liquor which was so completely devoid ( after the CaC03 treatment ) of anthocyan that its presence could be neglected .
When such a liquqr was freed from ether by distillation , it showed but the faintest pink coloration on acidification with cone .
HC1 , but , on addition of a small amount of Mg ribbon , the solution rapidly developed a fine red colour , as would be expected if a flavone derivative were present .
The main bulk of the liquors were deprived of their alcohol , hydrolysed by boiling with acid , and treated exactly as described above .
In this case , when the ether solution obtained was examined , it was found to contain only a flavone derivative which gave a yellow colour with dilute sodium carbonate solution , whilst , from the yellow aqueous layer so obtained , a small deposit of a yellow powder was obtained after acidification and standing .
Hence there is present in the Viola under investigation not only a myricetin glucoside , but also another yellow sap-pigment , which does not give a green coloration ( in sugar-free condition ) with dilute alkalis .
Having failed to detect the presence of myricetin in the liquors prepared in the above manner , the crude anthocyan precipitate from which they were obtained was examined , and it was very readily shown that , by the more complete precipitation of the anthocyan pigment by use of 3\#163 ; volumes of ether in place of the 2^ previously used , the myricetin glucoside had been thrown down with the anthocyan , from which it was readily recovered .
The crude , undried , anthocyan precipitate , obtained in the above way , was Dr. A , E. Everest .
dissolved in slightly warm acidified ( HOI ) ethyl alcohol , the solution filtered and the anthocyan reprecipitated by the addition of 2| times its volume of ether .
The liquors were decanted , filtered through a double filter , and treated exactly as before to show the presence of myricetin .
When the washed and filtered ether solution obtained in this case\#151 ; the final filtration of the washed ether extract , before testing for the colour reaction , was always included as a precaution against the possibility that a green colour may be formed as the result of some small amount of suspended matter being present\#151 ; was shaken with dilute aqueous sodium carbonate solution , a splendid clear deep green was obtained that passed through the colour changes described above .
In view of the fact that , as prepared in this case , the ether solution contained no waxes , and A. Gr .
Perkin has pointed out that the colour reactions described above can only be taken as a conclusive test for the presence of myricetin ( or less probably gossypetin ) if it is quite certain that it is an ether-soluble substance that produces them , despite the fact that the precautions taken above appear to preclude the presence of any ether-insoluble product when the tests were made , a bulk of the ether solution prepared as in the last case was evaporated to dryness , whereby a yellow residue was obtained which also dissolved in dilute alkalis to give the colour reactions already described\#151 ; green passing through bluer-green to brown-green , then orange-yellow . .
As a further precaution , some of the yellow residue was extracted with fresh ether , the ether extract filtered through double filter paper , and half the filtrate shaken with dilute sodium carbonate solution , when a fine green aqueous layer was obtained which passed through the same colour changes ; the other half of the ether extract was evaporated , and the residue treated with dilute sodium carbonate solution , when it also produced the same colour reaction .
Furthermore , the residue obtained from the bulk of ether solution originally evaporated when dissolved in acid ethyl alcohol and treated with Mg gave a clear , strong , red coloration .
' These observations leave one possibility to be removed , viz. , that the green coloration results from some product of decomposition of the anthocyan pigment during the hydrolysis , or from some unexplained retention of the anthocyan pigment , in colourless form , by the ether .
This possibility was removed by taking 0*1 grm. of the pure crystalline anthocyanin chloride ( violanin chloride)\#151 ; a quantity greatly in excess of that present in the ether-alcohol liquors\#151 ; dissolving it in alcohol , concentrating the solution under the conditions used for the extracts above mentioned , hydrolysing with boiling acid , then extracting the cooled product with ether , washing the ether extract , finally filtering the washed ether layer , all under exactly the same The Production of Anthocyanins and Anihocyanidins .
265 conditions as used in the examination of the yellow pigments .
The ether layer thus obtained , when shaken with dilute sodium carbonate solution , gave no trace of green coloration .
In the light of the results thus far obtained , the methods employed in these preliminary investigations are capable of considerable simplification , and by taking advantage of this it is hoped that the isolation of both myricetin and the sap pigment , which gives a yellow coloration with alkalis , will be possible in sufficient quantities for confirmation of their identities by means of derivatives .
It is doubtful whether the quantities present will allow of an insight into the nature of their glucosides being obtained , but an attempt will be made if it appears reasonable .
When time and circumstances permit , it is proposed to extend this work , in the first place , to various other Violas , then to Rosa gallica ( in which the anthocyanin is cyanin ) , the Pelargonium zondle ( in which the anthocyanin is pelargonin ) , and other flowers , with a view to a systematic investigation on the lines developed in the present paper , and in suitable cases the examination of the glucoside pigments with a view to discovering whether , where the anthocyan pigments are accompanied by corresponding flavone derivatives , the two classes of pigment occur attached to the same sugars .
Having proved that the anthocyanin described in the present paper is identical with Willstatter and Weil 's violanin , it is not the author 's intention to encroach upon the further work outlined by them ( .
cit. ) concerning the details of its constitution .
VOL. XC.\#151 ; B. Y

